One curious fact, already mentioned, is that the photospheric absorption spectrum is visible in the region up to 4200 Å. After this limit, the spectrum is dominated by emission lines. The lines which are visible correspond mostly to hydrogen and neutral helium, which can be used to classify the star as B2. Such a classification was already proposed by Swings & Struve (1940). Although this type corresponds to the presence of the lines observed, the object is definitely not a well behaved B2 object. In the first place the equivalent widths of the He I lines are too weak by a factor up to two, which according to the interpretation of Israelian et al. (1996) is due to emissions filling in the absorption lines. Now if filling-in exists, one would expect it to be present also at other lines, like those of Si II, C II and H. However the equivalent widths of these lines correspond to B2. In second place we do not see the CIII-O II lines in the interval 4070-86. And in last place we see the Ca II line 3933 which is characteristic of later stars (around B 5). If veiling is present, we should have an even stronger Ca II line, which would make the disagreement even worse.
As far as luminosity is concerned, the star has been classified as luminosity class IV by Burnichon et al. (1967) This is however not a spectrocopic classification, since the classical criteria involving He I lines (4121, 4144, 4009) and N II (3995) cannot be applied because the helium are affected by fill-in and the N II line is absent. Observe also that there is an emission line at 3993, which would affect anyway the N II line.
The classification by Burnichon et al. (1967) is based upon the Chalonge - Barbier spectrophotometric system, which uses the Balmer discontinuity. In that paper the authors find that the three indices of the system are sligthly variable, a fact which they attribute to the influence of the emission envelope.
In view of all this, it seems best to state cautiously that the object is probably an object lying on or near to the main sequence. From a purely spectroscopic viewpoint it can also be added that the star is certainly not a supergiant.